GenAIExamples/SearchQnA/docker_compose/intel/hpu/gaudi/compose.yaml

# Copyright (C) 2024 Intel Corporation
# SPDX-License-Identifier: Apache-2.0

services:
  tei-embedding-service:
    image: ghcr.io/huggingface/tei-gaudi:1.5.0
    entrypoint: /bin/sh -c "apt-get update && apt-get install -y curl && text-embeddings-router --json-output --model-id ${EMBEDDING_MODEL_ID} --auto-truncate"
    container_name: tei-embedding-gaudi-server
    ports:
      - "3001:80"
    volumes:
      - "${MODEL_CACHE:-./data}:/data"
    runtime: habana
    cap_add:
      - SYS_NICE
    ipc: host
    environment:
      no_proxy: ${no_proxy}
      http_proxy: ${http_proxy}
      https_proxy: ${https_proxy}
      HABANA_VISIBLE_DEVICES: all
      OMPI_MCA_btl_vader_single_copy_mechanism: none
      MAX_WARMUP_SEQUENCE_LENGTH: 512
      INIT_HCCL_ON_ACQUIRE: 0
      ENABLE_EXPERIMENTAL_FLAGS: true
      host_ip: ${host_ip}
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://$host_ip:3001/health || exit 1"]
      interval: 10s
      timeout: 10s
      retries: 60
  embedding:
    image: ${REGISTRY:-opea}/embedding:${TAG:-latest}
    container_name: embedding-gaudi-server
    depends_on:
      tei-embedding-service:
        condition: service_healthy
    ports:
      - "3002:6000"
    ipc: host
    environment:
      no_proxy: ${no_proxy}
      http_proxy: ${http_proxy}
      https_proxy: ${https_proxy}
      TEI_EMBEDDING_ENDPOINT: ${TEI_EMBEDDING_ENDPOINT}
      HF_TOKEN: ${HUGGINGFACEHUB_API_TOKEN}
      LOGFLAG: ${LOGFLAG}
    restart: unless-stopped
  web-retriever:
    image: ${REGISTRY:-opea}/web-retriever:${TAG:-latest}
    container_name: web-retriever-server
    ports:
      - "3003:7077"
    ipc: host
    environment:
      no_proxy: ${no_proxy}
      http_proxy: ${http_proxy}
      https_proxy: ${https_proxy}
      TEI_EMBEDDING_ENDPOINT: ${TEI_EMBEDDING_ENDPOINT}
      GOOGLE_API_KEY: ${GOOGLE_API_KEY}
      GOOGLE_CSE_ID: ${GOOGLE_CSE_ID}
    restart: unless-stopped
  tei-reranking-service:
    image: ghcr.io/huggingface/text-embeddings-inference:cpu-1.6
    entrypoint: /bin/sh -c "apt-get update && apt-get install -y curl && text-embeddings-router --json-output --model-id ${RERANK_MODEL_ID} --auto-truncate"
    container_name: tei-reranking-server
    ports:
      - "3004:80"
    volumes:
      - "${MODEL_CACHE:-./data}:/data"
    shm_size: 1g
    environment:
      no_proxy: ${no_proxy}
      http_proxy: ${http_proxy}
      https_proxy: ${https_proxy}
      host_ip: ${host_ip}
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://$host_ip:3004/health || exit 1"]
      interval: 10s
      timeout: 10s
      retries: 60
  reranking:
    image: ${REGISTRY:-opea}/reranking:${TAG:-latest}
    container_name: reranking-tei-gaudi-server
    depends_on:
      tei-reranking-service:
        condition: service_healthy
    ports:
      - "3005:8000"
    ipc: host
    environment:
      no_proxy: ${no_proxy}
      http_proxy: ${http_proxy}
      https_proxy: ${https_proxy}
      RERANK_TYPE: ${RERANK_TYPE}
      TEI_RERANKING_ENDPOINT: ${TEI_RERANKING_ENDPOINT}
      HUGGINGFACEHUB_API_TOKEN: ${HUGGINGFACEHUB_API_TOKEN}
      LOGFLAG: ${LOGFLAG}
    restart: unless-stopped
  tgi-service:
    image: ghcr.io/huggingface/tgi-gaudi:2.3.1
    container_name: tgi-gaudi-server
    ports:
      - "3006:80"
    volumes:
      - "${MODEL_CACHE:-./data}:/data"
    environment:
      no_proxy: ${no_proxy}
      http_proxy: ${http_proxy}
      https_proxy: ${https_proxy}
      HUGGING_FACE_HUB_TOKEN: ${HUGGINGFACEHUB_API_TOKEN}
      HF_HUB_DISABLE_PROGRESS_BARS: 1
      HF_HUB_ENABLE_HF_TRANSFER: 0
      HABANA_VISIBLE_DEVICES: all
      OMPI_MCA_btl_vader_single_copy_mechanism: none
      ENABLE_HPU_GRAPH: true
      LIMIT_HPU_GRAPH: true
      USE_FLASH_ATTENTION: true
      FLASH_ATTENTION_RECOMPUTE: true
    healthcheck:
      test: ["CMD-SHELL", "curl -f http://$host_ip:3006/health || exit 1"]
      interval: 10s
      timeout: 10s
      retries: 100
    runtime: habana
    cap_add:
      - SYS_NICE
    ipc: host
    command: --model-id ${LLM_MODEL_ID} --max-input-length 2048 --max-total-tokens 4096
  llm:
    image: ${REGISTRY:-opea}/llm-textgen:${TAG:-latest}
    container_name: llm-textgen-gaudi-server
    depends_on:
      tgi-service:
        condition: service_healthy
    ports:
      - "3007:9000"
    ipc: host
    environment:
      no_proxy: ${no_proxy}
      http_proxy: ${http_proxy}
      https_proxy: ${https_proxy}
      LLM_ENDPOINT: ${TGI_LLM_ENDPOINT}
      LLM_MODEL_ID: ${LLM_MODEL_ID}
      HUGGINGFACEHUB_API_TOKEN: ${HUGGINGFACEHUB_API_TOKEN}
      HF_HUB_DISABLE_PROGRESS_BARS: 1
      HF_HUB_ENABLE_HF_TRANSFER: 0
      LOGFLAG: ${LOGFLAG}
    restart: unless-stopped
  searchqna-gaudi-backend-server:
    image: ${REGISTRY:-opea}/searchqna:${TAG:-latest}
    container_name: searchqna-gaudi-backend-server
    depends_on:
      - tei-embedding-service
      - embedding
      - web-retriever
      - tei-reranking-service
      - reranking
      - tgi-service
      - llm
    ports:
      - "3008:8888"
    environment:
      - no_proxy=${no_proxy}
      - https_proxy=${https_proxy}
      - http_proxy=${http_proxy}
      - MEGA_SERVICE_HOST_IP=${MEGA_SERVICE_HOST_IP}
      - EMBEDDING_SERVICE_HOST_IP=${EMBEDDING_SERVICE_HOST_IP}
      - WEB_RETRIEVER_SERVICE_HOST_IP=${WEB_RETRIEVER_SERVICE_HOST_IP}
      - RERANK_SERVICE_HOST_IP=${RERANK_SERVICE_HOST_IP}
      - LLM_SERVICE_HOST_IP=${LLM_SERVICE_HOST_IP}
      - EMBEDDING_SERVICE_PORT=${EMBEDDING_SERVICE_PORT}
      - WEB_RETRIEVER_SERVICE_PORT=${WEB_RETRIEVER_SERVICE_PORT}
      - RERANK_SERVICE_PORT=${RERANK_SERVICE_PORT}
      - LLM_SERVICE_PORT=${LLM_SERVICE_PORT}
      - LOGFLAG=${LOGFLAG}
    ipc: host
    restart: always
  searchqna-gaudi-ui-server:
    image: ${REGISTRY:-opea}/searchqna-ui:${TAG:-latest}
    container_name: searchqna-gaudi-ui-server
    depends_on:
      - searchqna-gaudi-backend-server
    ports:
      - "5173:5173"
    environment:
      - no_proxy=${no_proxy}
      - https_proxy=${https_proxy}
      - http_proxy=${http_proxy}
      - BACKEND_BASE_URL=${BACKEND_SERVICE_ENDPOINT}
    ipc: host
    restart: always

networks:
  default:
    driver: bridge